What a CRM implementation actually costs, and what is in it

Telling detail: almost no CRM integrator publishes a price for implementation. Everywhere it is "request a quote" and "we'll price it individually". So let us start with what can be calculated exactly.

The worked example below uses KeyCRM, because its pricing is genuinely public and denominated in dollars, which makes it checkable. It is strongest for stores in Ukraine and the CIS; the arithmetic and the structure of the decision apply to any CRM you are weighing up.

What the software itself costs

What	How much
Base plan	$19 a month
What that includes	200 orders, 2,000 leads, 20,000 chat messages, 10 GB, unlimited users
Over the limit	+$9 per 200 orders
Extra storage	+$9 per 10 GB
Over the lead / message limit	+$9 per 2,000 leads or 20,000 messages
Marketplace module	+$19 a month, unlimited connections
Dedicated Telegram number	+$15 a month
WhatsApp number	+$20 a month
Dedicated Viber number	+$25 a month
Catalog products	unlimited
Trial	up to 30 days

Let us price a real store. 400 orders a month, two marketplaces, a dedicated messenger number for customers:

Base	$19
Second hundred orders over the limit	$9
Marketplace module	$19
Messenger number	$25
Total	$72 a month

Note the last line: messengers are a separate cost, and they are the one people forget when they calculate "a CRM for nineteen dollars". If you need your own numbers on Telegram, WhatsApp or Viber, that adds more than the system itself costs.

An important detail in the other direction: users are not charged for, and catalog products are unlimited. That is unusual — most competitors bill per manager, so the invoice grows linearly as the team does.

What integrators actually charge for

Since nobody publishes prices, judge it by scope of work. A full implementation is:

• Describing the process before automating it. The most important stage and the one skipped most often. Move a chaotic process into a CRM and you get chaos with a dashboard.
• Configuring the pipeline and statuses around your real order journey, not a template one.
• Connecting the sources: the site, Instagram and DMs, Telegram, marketplaces, classifieds, telephony.
• Shipping and payments: carriers with automatic labels and status updates, payment providers, receipts.
• Products and stock: catalog sync, so you do not sell what you do not have.
• Roles and permissions — so a manager sees their own work rather than the entire customer base.
• Reports built around the questions you genuinely ask.
• Training the team and supporting the first few weeks.

By scope that is anywhere from a few days to two or three weeks of work. Price from there: in my own pricing this is part of the "Setup and automation" rung, from $5,000 — and a smaller scope comes out cheaper. I publish that figure precisely because nobody else does.

How long it really takes

A modern CRM will fairly claim speed: a basic connection genuinely takes a day, and managers will immediately start handling orders from every source in one window. But "a day" describes the connection, not the implementation.

The difference is that in a day you get a tool, and in two weeks you get a process. Between those two states sits work that never appears in a demo: agreeing the statuses, settling who is responsible for what, rewriting reply templates, and getting the team to stop keeping a parallel record in a notebook.

When you do not need a CRM

Against my own interest: if you have under 50 orders a month and one source, you do not need a CRM. A spreadsheet will cope, and the $19 and two weeks are better spent elsewhere.

A CRM becomes mandatory when two conditions coincide: several order sources (site plus DMs plus a marketplace) and more than one person handling them. It is at the intersection of those two that orders go missing.

How to tell you are already losing money without one

Test yourself on three questions:

1. Can you say within a minute how many orders are in progress right now and what stage each is at?
2. Has an order from DMs been lost in the last month, or answered a full day late?
3. Do you know what percentage of cash-on-delivery orders go uncollected — broken down by channel?

Three "no"s and a CRM pays for itself within a couple of months on the orders you stop losing alone.

The main implementation mistake

Automating what should have been abolished. Very often the order process contains steps that exist "because that is how it ended up": a double confirmation, a manual copy into another spreadsheet, an approval nobody reads. Move that into a CRM and you have locked the unnecessary work in permanently — and paid for the privilege.

Which is why a sound implementation starts not with settings but with the question "what here could we simply stop doing?"